Fallacy 3: Bandwidth is Infinite
You'll learn why assuming unlimited network capacity is dangerous and how to design systems that respect bandwidth constraints.
The Core Misconception
Many developers assume that network bandwidth is essentially unlimited—that you can send as much data as you want without consequences. In reality, network capacity is a precious, finite resource. Even modern high-speed connections have limits, and those limits become painfully obvious when multiple services compete for bandwidth or when clients connect from mobile networks, rural areas, or congested corporate networks.
The problem: Large payloads, uncompressed data, and careless data transfer patterns quickly saturate available bandwidth, creating bottlenecks that degrade performance system-wide.
Why This Matters
Think of bandwidth like a highway. A single car travels smoothly, but when hundreds of trucks carrying oversized loads hit the road simultaneously, traffic grinds to a halt. Your "oversized loads" might be:
- Sending entire database records when only a few fields are needed
- Returning uncompressed JSON responses
- Transferring high-resolution images without optimization
- Polling frequently with large payloads instead of using efficient protocols
- Not paginating large result sets
Respecting Bandwidth Constraints
Be mindful of payload size: Only send data that's actually needed. Use field selection, pagination, and filtering at the API level.
Use compression: Enable gzip or Brotli compression for HTTP responses. The CPU cost of compression is usually far less than the network cost of larger payloads.
Consider data transfer costs: In cloud environments, bandwidth often has direct monetary costs—both for data transfer and for the time clients spend waiting.
Monitor and measure: Track payload sizes and bandwidth usage to identify optimization opportunities before they become bottlenecks.
Key Takeaway: Network bandwidth is limited and costly; design systems that minimize payload size through compression, selective data transfer, and efficient protocols to prevent performance bottlenecks and reduce operational costs.